
Agentlas
Run a team of AI agents from one Home. Open source.
3 followers
Run a team of AI agents from one Home. Open source.
3 followers
Agentlas 2.0 puts all your AI agents in one Home. Group rooms have a lead: @ anyone, or watch agents hand work to each other in the open. Run chats, rooms and work sessions side by side in up to 4 panes. Use the plan you already pay for (Claude Code, Codex, Cursor, Copilot CLI) or local models, so no extra bill and keys stay in your keychain. Download agents from the Hub and make them yours. Free to start, open source (Apache-2.0), on macOS, Windows and Linux.
This is the 2nd launch from Agentlas. View more
Agentlas
Launching today
Chats, group rooms and work sessions in one Home, side by side. Rooms have a lead, agents hand work to each other in the open, and it all runs on the plans and keys you already have. Local and open source.






Launch Team

CloudflareYour next app belongs on Cloudflare Workers. Deploy for free
Promoted

Hi Product Hunt,
Two things happened since our first launch here: Agentlas stopped being "describe a worker, get an agent" and became a place where your agents live. 2.0 is that place.
What is new:
One Home. Projects, agent chats, group rooms and work sessions sit in one tree. Open up to four of them side by side.
Rooms with a lead. A message without an @ goes to the lead. @ anyone to ask directly. Agents can @ each other, so you watch the hand-offs instead of trusting a hidden coordinator.
Your plan, your keys. Use Claude Code, Codex, Cursor or Copilot CLI on the subscription you already pay for, or run local models. Agentlas never proxies the model call and keys stay in your OS keychain.
Open source (Apache-2.0), on macOS, Windows and Linux.
I would love to know: what is the first team of agents you would put in a room? Criticism and GitHub issues are welcome.